Hazardous and binge drinking have been associated with increased risky sexual behavior, poor adherence to antiretroviral therapy (ARVs) and toxicity from ARVs among those with HIV infection. As such, hazardous and binge drinking (score of e3 on the Alcohol Use Disorders Test - Consumption (AUDIT-C), or e6 drinks per occasion at least monthly) have a major impact on HIV transmission and disease progression. Prevalence rates of hazardous drinking are particularly high among HIV (53%) and general medicine (68%) outpatients in western Kenya, in part due to the wide availability of potent traditional brew. Growing evidence suggests that heavy drinking is an obstacle to successful sexual risk reduction approaches. This team recently completed a Stage 1 trial of a group cognitive-behavioral therapy (CBT) intervention delivered by paraprofessionals to reduce alcohol use among 74 HIV-infected Kenyans (R21AA017884). Results demonstrated feasibility, acceptability and estimated a large treatment effect at post-treatment. We now propose a large Stage 2 efficacy trial of the CBT alcohol intervention. Our goal is to evaluate the efficacy of the Reduce Alcohol First in Kenya Intervention (RAFIKI) in its ability to reduce alcohol use within a larger Stage 2 trial that includes an active control and a longer follow-up period. Rafiki means friend in Kiswahili, the national language of Kenya. The trial will be conducted by the Kenya Health Behavior Study (KHBS) team, an experienced group of Kenyan and U.S. behavioral scientists, physicians, substance users in recovery and persons infected with HIV. KHBS expands on well-established ties between the Academic Model for Providing Access to Healthcare (AMPATH) and the Brown University Medical School, which has been an active AMPATH partner since 1997. AMPATH currently treats more than 75,000 HIV-infected patients in 25 clinics in western Kenya. As part of these efforts, we have ready locally adapted training, treatment and fidelity rating manuals, a successful paraprofessional intervention delivery model, clinical assessment tools, and an experienced local team of trainers and supervisors. Our specific aims are to: 1) To examine the efficacy of a same-sex group CBT intervention in a randomized clinical trial of 336 HIV-infected Kenyan outpatients who report hazardous or binge drinking, when compared to a time-matched group Healthy Lifestyles education intervention; 2) To conduct analyses to examine the mechanisms of intervention effects; and 3) To conduct exploratory analyses to examine the relationship between intervention condition and sexual risk behaviors. Completion of our objectives, which are consistent with NIAAA's mission to reduce both alcohol use and HIV risk in vulnerable populations, will provide a robust test of efficacy of the paraprofessionally led group CBT and potentially provide a sustainable and transportable intervention for other settings in sub-Saharan Africa.